Bedform characterization in river channel through 2D spectral analysis

advertisement
Marine and River Dune Dynamics – MARID IV – 15 & 16 April 2013 - Bruges, Belgium
Bedform characterization in river channel through 2D spectral
analysis
A. Lisimenka(1), S. Rudowski(1)
1. Maritime Institute in Gdansk, Gdansk, Poland - sasha@im.gda.pl, starud@im.gda.pl
Abstract
During last decades a multibeam echosounder has become a widely used tool for a high resolution, rapid and, first
of all, cost-effective mapping of submerged environments. At the present time, wealthy literature related to the
bedform characterization based on the multibeam swath bathymetry systems data collected in different aquatic
areas all over the World exist. In this paper, results of the recent bathymetry investigations in the main Vistula river
mouth − the Cross-cut artificial channel, which were performed several times during 2009-2012 years are
presented. The comprehensive riverbed morphology of the channel is characterized by the set of the different kinds
and patterns of sandy bedforms. A detailed analysis of the bottom roughnesses in conjunction with a quantitative
parameterization of the primary geometrical characteristics of the bedforms based on two-dimensional spectral
analysis (2D FFT) are executed. The obtained results indicates significant changes of the morphology and bedform
patterns which are very important in order to estimate and verify the bedload sediment transport phenomena and
hydrodynamic processes connected with. Moreover, such information is required for predicting of the flood
possibility related to riverine and sea storm surges in the area of interest.
1. INTRODUCTION
Different type of bottom surface roughnesses, such
as ripples, megaripples, sandwaves and dunes, are
a fundamental seafloor features which reflect and
at the same time impact on a variety of physical
phenomena including sediment transport and
hydrodynamic processes. Characterization of this
rhythmic bedforms essential parameters, their
spatial distribution, dimension and orientation, as
well as its migration and evolution, is with no
doubts required for understanding and quantifying
the influence of sediment topography on current
state of a aquatic system and can be useful for
predicting of its natural and anthropogenic
changes.
The oceanographers' community manifests
growing interest in a quantitative echosounders
data analysis in context of the sediment transport
estimation since the sonar systems of sea bottom
visualization have appeared (among others,
Kenyon, 1970, Kenyon et al., 1981). But just in
last years, development of the sonar technologies,
computer techniques and signal processing
methods has led to more fully exploitation of an
information obtained by also in case of multibeam
177
echosounder systems (MBES). It has found a
reflect in growing numbers of the scientific works
encapsulating this topic.
Some classification schemes of the various type of
the bedforms were proposed (Ashley, 1990,
Mitchell & Hughes Clarke, 1994, Knaapen et al.,
2001, Tegowski et al., 2011), models of wind
waves and currents interaction with seabottom
were created (Davis et al., 2004).
Using of a high-frequency MBES gave a
possibility to determine a total volumetric
sediment transport in a tidal channel (Bartholomä
et al., 2004) and a sand dunes crests migration
over a tidal period in a shallow water (Knaapen et
al., 2005).
An attempt of finding a relationship between
precise morphology, three-dimensional flow
structure and sediment transport over a large sand
dunes was presented in Parsons et al. (2005, 2006).
Detailed information about bottom topography
changes together with space distribution of the
flow speed and grain size analysis of the bed
material allowed the direct determination of bed
resistance during flood on the Rhine river (Julien
et al., 2002).
Marine and River Dune Dynamics – MARID IV – 15 & 16 April 2013 - Bruges, Belgium
High variability of the main bedform structures
parameters in space and in time requires an
appropriate mathematic tool for its determination.
In context of scientific problem presented in this
paper, results of works of many investigators based
on algorithms of two-dimensional Fast Fourier
Transform (2D FFT) spectral analysis have
indicated a possibility of detailed quantification of
the primary bedforms characteristics, determine
their space form and dominant surface roughnesses
(among others, Percival & Walden, 1993, Lyons et
al., 2002, Briggs et al., 2005, Cazenave et al.,
2008, Lefebvre et al., 2011, Lefebvre & Lyons,
2011), what is a main term for verification of the
bedload sediment transport and hydrodynamic
processes connected with.
2.
MEASUREMENT AREA
Investigations were carried out in the main Vistula
river mouth located in Polish coastal waters (Fig.
1). The Vistula is the longest river in Poland and
plays a dominant role as regards both the
catchment area size and the amount of discharged
water flowing into the Gulf of Gdansk.
Generally, a hydrological conditions in the main
Vistula river mouth area, where a river waters mix
with sea waters from the Gulf of Gdansk, are
formed through interaction of the processes from
this two environments. This riverine and marine
processes influence superbly on not only current
hydrological regime in the area, but seriosly affect
on the morphology evolution of this part of the
river, coast and neighbouring marine waters in
longterm period of time.
Scale of the hydrological conditions related to the
magnitude and variation of the water flows during
2009-2012 hydrological years is presented in
Fig.2. Average daily water discharge data1 were
obtained from the gauging section located in
Tczew, in range of 31.2km upstream the Vistula
channel mouth. This gauging section is the last
representative place situated in the lower reach of
the Vistula river, which covers 99,92% of the
catchment area (Augustowski, 1982), and,
moreover, is not subjected to the impact of the
backwater phenomena connected with the sea
storm surge (mainly occurs when strong winds
blow from north and west directions during long
period of time).
°
6000
3
Water Discharge [ m /s ]
55.0 N
B a l t i c
S e a
2009
2010
5000
2011
4000
2012
3000
2000
1000
Gulf of Gdansk
XI
°
54.5 N
XII
I
II
III
IV
V
VI
VII
VIII
IX
X
Month
Figure 2. Average daily water discharge during
2009-2012 hydrological years.
Measurement
area
POLAND
°
18.5 E
19.0° E
19.5° E
Figure 1. Location of the measurement area.
The main Vistula river mouth is the Cross-cut
artificial channel with a ca 3000m length, 400m
width and up to 10m depth. More than 95% of
total Vistula waters outflow into the sea through
this channel with the average annual water
discharge more than 1000m3/s (Cyberski, 1997).
178
In period of time between two measurement
campaigns (see Section 3.1), a considerable
variations in water discharge were taken place.
Several phenomena with extremely high
magnitudes connected with a precipitation floods
(up to 6360m3/s in May 2011) and a snow-melt
floods (up to 4000m3/s in January 2011) were
observed and with no doubts have significantly
1
hydrological data source − internet information service
of the Institute of Meteorology and Water Management,
www.pogodynka.pl
Marine and River Dune Dynamics – MARID IV – 15 & 16 April 2013 - Bruges, Belgium
influenced on morphological character of the
bottom as well as sediment mass transport in the
area of interest.
3.
3.1.
METHODS
Data acquisition
Bathymetry measurements were carried out several
times in 2009-2012 years on the board of the
hydrographic boat Imoros 2. In the paper, results
of two measurement campaigns which were
performed in 20-23 August 2009 and 4-6 June
2012 are presented. An area of the interest was
mapped using boat mounted high-resolution Reson
Seabat 8101 (before 2012 year) and 7101
multibeam echosounder systems (MBES). The
positioning system DGPS RTK Trimble SPS 851
together with Ixsea Hydrins inertial navigation
system were integrated with MBES using the
QINSy data acquisition software package. During
of the each hydroacoustic surveys, the vertical
sound speed profiles were measured using the
portable model Reson Sound Velocity Probe
SVP15.
3.2.
Postprocessing
The bathymetric data were gridded with a cell size
of 0.25m. The whole dataset was subsequently
divided into the sub-domains using by a sliding
square windows of 50m side with a 75%
overlapping between an adjacent windows. Within
a particular window, the mean depth was
subtracted from each data sample and then a
detrending procedure was applied. In order to
minimize the spectral leakage due to the finite
length of the data set, the data in each window
were pre-multiplying by single orthogonal taper
belongs to a family of functions known as discrete
prolate spheroidal sequences (DPSS). Next, the
spectral analysis based on the 2D Fast Fourier
Transform (2D FFT) algorithm was executed:
S( fx , f y ) =
+∞+∞
∫ ∫ s ( x, y ) e
− 2πi ( f x x + f y y )
dxdy
(1)
− ∞− ∞
where s(x,y) = spatial signal in the x and y
directions, S(fx, fy) = two-dimensional amplitude
spectrum referred to the appropriate spatial
frequencies fx and fy.
In order to exclude appearance of false
components in spectra associated with the border
of the data coverage effect, the 2D FFT was
179
calculated only in sub-domains with a full data
coverage (completely wet areas).
Then, the power spectrum density was calculated
and direction and spatial frequency of the most
energetic peak were determined.
As it was also noted in Lefebvre et al. (2011),
although the data were detrended before being
spectrally transformed, low frequency components
were still present and masked the higher
frequencies that are of interest. So, frequencies
corresponding to wavlengths larger than 25m (half
a square window geometrical dimension) were
filtered out.
4.
RESULTS
The comprehensive morphological character of the
bottom is presented (Fig.3). The bottom relief are
consist of two biggest forms: sand bar and trough,
located along the border of the channel. In August
2009 (Fig. 3a), this main forms occupied almost all
of the channel area. But at the end part, the trough
is closed by the sandbar and nearly filled the
channel mouth, in spite of the left side where an
erosional depression is situated. In June 2012 (Fig.
3b), there are significant changes in bottom
morphology. The sandbar is become longer,
completely covered the erosional depression. By
another hand, the trough end is become good
formed completely opened the mouth. Moreover, a
set of the different bedforms as sandwaves with
superimposed megaripples and ripples with
various height, wavelength and orientation are
observed.
The calculated bedform orientation and
wavelength based on the results of the twodimensional spectral analysis in particular square
windows, depicted by black vectors and
superimposed on the part of the appropriate bottom
relief (depicted by dashed rectangulars in Fig.3)
are shown in more detailed scale in Fig.4. The
different pattern of the bedform crests and
significant changes of the their spatial frequency
and orienation as well as distribution are distinctly
visible.
5.
CONCLUSIONS
The obtained results indicates that the application
of the 2D spectral analysis is a good tool for
Marine and River Dune Dynamics – MARID IV – 15 & 16 April 2013 - Bruges, Belgium
studies of the bedload sediment transport
phenomena, roughnesses character of the bottom
and hydrodynamic processes connected with. It
should be require for monitoring, modeling and
prognosis of the bedload transport and flows,
especially related to flood possibility (riverine
and/or sea storm surge). The data are also needed
to prepare an emergency warning against extreme
hydrological conditions for local society by
appropriate services.
6.
ACKNOWLEDGMENT
The author is grateful to the National Centre
for Research and Development for funding
this work by grant No PBS1/A2/3/2012.
7.
REFERENCES
Ashley, G. M. 1990. Classification of large-scale
subaqueous bedforms; a new look at and old
problem. J. Sediment. Res. 60(1): 161-172.
Augustowski, B. (ed.). 1982. Dolina dolnej Wisły.
Ossolineum, Wrocław (in polish).
Bartholomä, A. & Ernstsen, V.B. & Flemming, B.W. &
Bartholdy, J. 2004. Bedform dynamics and net
sediment transport paths over a flood-ebb tidal
cycle in the Grådyb channel (Denmark),
determined
by
high-resolution
multibeam
echosounding. Danish J. of Geography 104(1): 4555.
Briggs, K.B. & Lyons, A.P. & Pouliquen, E. & Mayer,
L.A. & Richardson, M.D. 2005. Seafloor
roughness, sediment grain size and temporal
variability. Proc. of Intern. Conf. and Exchib. on
UAM, Heraklion, Crete, Greece, pp. 29-37.
Cazenave, P. W. & Lambkin, D.O. & Dix, J.K. 2008.
Quantitative bedform analysis using decimetre
resolution swath bathymetry. Caris 2008
Conference Proc., 22-26.09.2008, Bath, UK, 12 p.
Cyberski, J. 1997. Riverine outflow into the Gulf of
Gdańsk. Oceanol. Stud. 26(4): 65–75.
Davis, J.P. & Walker, D.J. & Townsend, M. & Young,
I.R. 2004. Wave-formed sediment ripples:
Transient analysis of ripple spectral development.
J. Geoph. Res. 109 (C07020): 15 pp.
doi:10.1029/2004JC002307
Julien, P. Y. & Asce, M. & Klaassen, G. J. & Ten
Brinke, W. B. M. & Wilbers, A. W. E. 2002. Case
180
Study: Bed Resistance of Rhine River during 1998
Flood. J. Hydraul. Eng. 128: 1042-1050.
Kenyon, N.H. 1970. Sand ribbons of European tidal
seas. Marine Geology 9: 25-39.
Kenyon, N.H. & Belderson, R.H. & Stride, A.H. &
Johnson, M.A. 1981. Offshore tidal sand banks as
indicators of net sand transport and as potential
deposits. Spec. Public. of the Intern. Assoc. of
Sedimentologists. 5: 257-268.
Knaapen, M.A.F. & Hulscher, S.J.M.H. & de Vriend,
H.J. 2001. A new type of sea bed waves.
Geophysical Research Letters 28(7): 1323-1326.
Knaapen, M.A.F. & van Bergen Henegouw, C.N. & Hu,
Y.Y. 2005. Quantifying bedforms migration using
multi-beam sonar. Geo-Marine Letters 25: 306-314.
Lefebvre, A. & Ernstsen, V.B. & Winter, C. 2011.
Bedform characterization through 2D spectral
analysis. J. Coast. Res. Spec. Issue 57: 781-785.
Lefebvre, A. & Lyons, A.P. 2011 Quantification of
roughness for seabed characterization. Proc. of 4th
Intern. Conf. and Exhib. on UAM, Kos Island,
Greece, 20-24 June 2011, 8 p.
Lyons, A.P. & Fox, W.L.J. & Hasiotis, T. & Pouliquen
E. 2002. Characterization of the two-dimensional
roughness of wave-rippled sea floors using digital
photogrammetry. IEEE J. Ocean. Eng. 27(3): 515524.
Mitchell, N.C. & Hughes Clarke, J.E. 1994.
Classification of seafloor geology using multibeam
sonar data from the scotian shelf. Marine Geology.
121: 143-160.
Parsons, D. R. & Best, J. L. & Orfeo, O. & Hardy, R. J.
& Kostaschuk, R. & Lane, S. N. 2005. Morphology
and flow fields of three-dimensional dunes, Rio
Paraná, Argentina: Results from simultaneous
multibeam echosounding and acoustic Doppler
current profiling. J. Geophys. Res. 110 (F04S03): 9
pp.
Parsons, D. R. & Best, J. L. & Lane, S. N. & Hardy, R.
J. & Kostaschuk, R. & Shugar, D. & Orfeo, O.
2006. Morphology, flow and sediment transport
over a natural 3D dune field: Rio Paraná,
Argentina. River Flow 2006 vols 1 and 2: 9971004.
Percival, D.B. & Walden, A.T. 1993. Spectral analysis
for physical application. Cambridge University
Press, 616 p.
Tegowski, J. & Nowak, J. & Moskalik, M. & Szefler,
K. 2011. Seabed classification from multibeam
echosounder backscatter data using wavelet
transformation and neural network approach. Proc.
of 4th Intern. Conf. and Exhib. on UAM, Kos
Island, Greece, 20-24 June 2011, 1257-1264.
Marine and River Dune Dynamics – MARID IV – 15 & 16 April 2013 - Bruges, Belgium
(a)
(b)
Figure 3. Bathymetry images of the Vistula river channel on the base of measurements performed
in August 2009 (a) and June 2012 (b). Coordinates in WGS-84. Colorbar scales in meters.
(a)
(b)
Figure 4. Calculated bedform orientation and wavelength depicted in the black vectors superimposed on the
bottom relief of the river channel (part of the Fig.3 in bigger scale) for August 2009 (a) and June 2012 (b)
respectively. Length of the black vectors is proportional on bedform wavelength. Coordinates in WGS-84.
Colorbar scales in meters.
181
Download